188 ALASKAN GLACIER STUDIES
such as to give rise to the deepest possible breaking. We, therefore, have observational data for tracing the life history of a system of profound crevasses; and the result is, what is well known from the study of other crevasses more normally produced,—namely, that they are mere surface fissures in the very upper parts of the glacier. One or two hundred feet of ablation lowers the ice surface to their bottoms, and below that the ice is unbroken. Yet surely the lower ice was subjected to a profound straining of such abrupt nature that it should have responded by fissuring if its condition had permitted. Ice in a state of plasticity such as the theory of viscous flowage demands* could not be fissured by even the abrupt and powerful strains which accompanied the advance of the glaciers; but ice in a condition demanded by theories not involving a state of plasticity would be expected to yield to such strains by fissuring.
We cannot observe the condition of glacier ice in motion, and experimentation haa so far not even approximately duplicated the conditions existing in the interior and bottoms of glaciers. These facts account in part for the failure of inferences from theoretical considerations of the problem to bring forth results which are convincing to all. In these studies of the problem, known facts have been considered in the development of theory, but the facts are partly unknown, and there is necessity of assumption which if incorrect removes the foundation of the theory. In all of the theories of glacial motion there are assumptions upon which there is not general agreement,—hence the failure of even the most elaborately argued theories to receive universal acceptance. The theory of viscous flowage has lost favor with some students of the subject of glacial motion by reason of the fact that glacier ice is a crystalline rock; and the assumption has been made that this condition and the supposed inability to hold stones in the bottom and perform glacial erosion are incompatible with viscosity. Partly upon the basis of these assumptions several elaborate hypotheses have been put forth as a substitute for the theory of viscous flowage, which has numerous adherents.
Still maintaining our determination not to enter into a discussion of the literature on the physics of glacier motion, and confining ourselves primarily to a statement of facts of observation and the conclusions to which they lead, we will state some facts observed in the Yakutat Bay region which we believe have a direct bearing on the problem of the condition of glacier ice. While there, as in other regions, we cannot make direct observations on the condition of a large glacier from top to bottom, we nevertheless have samples of the glacier ice, from different parts of the glacier, supplied to us for observation in the form of innumerable icebergs, floating in the waters of the fiord and stranded on its beaches. From a study of these and of the clear glacier surfaces, sides of crevasses, and ice cliffs of the glaciers we can form some idea of the nature of the glacier from top to bottom.
The largest of the glaciers, the Hubbard, presents an ice cliff of marble whiteness some three hundred feet high, the upper half more or less profoundly crevassed. What lies beneath the water level is hidden, but the appearance of icebergs black with debris, and the abundance of debris-charged bergs in the waters of the fiord lead us to conclude that the very lowest layers are heavily burdened with morainic materials. Among the thousands of icebergs floating in the fiord and stranded on its shores about two-thirds are whitish and one-third are glassy, as shown by repeated observations at favorable localities (PL LXXIX, B). Some large bergs are entirely glassy, some entirely whitish, and some have bands of white and glassy ice, the proportions varying in dif-